A Computationally Efficient and Aliasing-Free Broadband DOA Estimation for Arbitrary Sparse Linear Arrays

Spatial aliasing is an inevitable issue for sparse linear arrays with the intersensory spacing larger than the half wavelength of the incident signals. Here, a computationally efficient and aliasing-free broadband DOA estimation algorithm is proposed. The array output is first multiplied by a designed diagonal matrix. In this manner, the frequency of the broadband signals is shifted to a given frequency which approximatively satisfies the spatial Nyquist sampling theorem. Thus, the spatial aliasing is avoided. Then the proposed method is demonstrated to be applicable to arbitrary sparse linear arrays. Simulation results are included to show the anti aliasing property and the computational efficiency of the proposed method.